"Save Link" of the current page performs Snapshot instead

"Save Link" of the current page performs Snapshot instead - and adds the whole document into ZOTERO base. Of course, in Preferences "Automatically take snapshots..." position is cleared. Frankly speaking, before trying ZOTERO I switched this position on. But when I found that it saves documents instead of just links to them, I cleared it and reloaded the Firefox. It was of no help. I've got ZOTERO v. 1.0.0rc3.
  • That pref is unrelated to the Save Link button, and I haven't seen anything like what you're reporting, but upgrade to 1.0.0rc5 from the home page and let us know if you're still having trouble.
  • Hi, Dan,
    Thank you for your prompt attention. I have upgraded my ZOTERO to 1.0.0rc5, removed the old data base, started the brand new one. Came to the http://fictionbook.ru/author/nikonov_aleksandr_petrovich/konec_feminizma_chem_jenshina_otlichaets/nikonov_konec_feminizma_chem_jenshina_otlichaets.html site. Then made a deep breath. My base at that moment was 757,760 bytes. Then I tried to save the link to the named page. It took considerable time, and then my ZOTERO data base became 3,188,736 bytes. I deleted this huge "link", returned to the starting DB size and tried to take the snapshot of that same page. The thing went into the endless loop, then it put the philosophical question - Stop or continue this script? Two times I let her continue, then stopped. She displayed an error and asked me to reload my Firefox. The DB was at its initial stage.
    By the way, I should warn you that all my life I was the luckiest user who stumbled upon all possible system bugs, while all others worked happily noticing nothing.
  • edited October 25, 2007
    Saving the snapshot took a long time because Zotero automatically indexes imported text, and you tried to add an entire book in Russian (not sure but perhaps non-Latin characters take longer?). If you had allowed the script to continue, it would have indexed the entire book eventually.

    To disable the warning message, go to about:config in the location bar and set dom.max_chrome_script_run_time to 0.
  • edited October 25, 2007
    Yessss... of course... But I am not very anxious about importing in the DB snapshots of several megabytes length. It was done only in the process of experimenting around saving links/snapshots. By the way, Russian symbols in the encoding used take exactly the same byte per symbol as Latin ones...
    Now my primary issue is about operation of copying links into ZOTERO DB. If it fails with this basic task - I will never use it in my practice, and, besides, how could it happen, that no one else did not notice it? Or it depends on some unclear factors in OS? How about software reliability then?
  • Saving the drowning people is their own responsibility, as Russians say: "Спасение утопающих - дело рук самих утопающих". My personal investigation showed that both operations: saving link and saving a snapshot give exactly the same result: they both save a snapshot. Initial ZOTERO DB is 780,288 bytes. Linking or snapshotting http://www.fictionbook.ru/author/alman_brayian_m/samogipnoz/alman_samogipnoz.html site brings its weight to 2,643,968 bytes . Can it be of any help for support guys?
  • Snapshots are stored outside the database as HTML/images. However, the page content of both snapshots and links are passed through the full-text indexer, which is why they increase the DB size the same amount.

    It might make more sense for links not to be indexed.

    Also, just as a side note, the encoding of the page doesn't matter. The content is converted to UTF-8 for insertion into the database, so the Russian characters are stored in two bytes rather than one, which might make storing Russian a bit slower (though not as slow as Asian languages).
  • edited October 26, 2007
    Thank you, Dan, - I see now what is going on. Is it explained in the documentation? Really, I was lacking this knowledge. By the way, you use the folder named "storage"; I spent some time searching for it... Maybe, it would be better to let user specify its name and exact position in the file structure? Considerations of always shortage of space... And, of course, lack of understanding concerning what and when is indexed and how to manage the whole process.
    In any case, the system is interesting... Thank you for your time!
  • You can specify the location of the data directory (which includes the database and the storage directory) in the Zotero prefs, and in the latest version there's also a button to reveal the directory in Explorer/Finder.

This is an old discussion that has not been active in a long time. Instead of commenting here, you should start a new discussion. If you think the content of this discussion is still relevant, you can link to it from your new discussion.